Abstract:There are more than a trillion sensors in the world today and according to some estimates there will be about 50 trillion cameras worldwide within the next 5 years, all collecting data either sporadically or around the clock. With such explosive growth of available data and computing resources, recent advances in machine learning and data analytics have yielded transformative results across diverse scientific disciplines, including image recognition, natural language processing, cognitive science, and genomics. However, in many engineering applications, quality and error-free data is not easy to obtain, e.g., for system dynamics characterized by bifurcations and instabilities, hysteresis, delayed responses, and often irreversible responses. Admittedly, as in all everyday applications, in engineering problems, the volume of data has increased substantially compared to even a decade ago but analyzing big data is expensive and time-consuming. Data-driven methods, which have been enabled in the past decade by the availability of sensors, data storage, and computational resources, are taking center stage across many disciplines (physical and information) of science. We now have highly scalable solutions for problems in object detection and recognition, machine translation, text-to-speech conversion, recommender systems, and information retrieval. All of these solutions attain state-of-the-art performance when trained with large amounts of data. However, purely data-driven approaches for machine learning present difficulties when the data is scarce and of variable fidelity relative to the complexity of the system. The vast majority of state-of-the art machine learning techniques (e.g., deep neural nets, convnets, RNNs, etc.) are lacking robustness and fail to provide any guarantees of convergence or quantify the error/uncertainty associated with their predictions. Hence, the ability to learn in a sample-efficient manner is a necessity in these data-limited domains. Less well understood is how to leverage the underlying physical laws and/or governing equations to extract patterns from small data generated from highly complex systems.

Computer simulations in science and engineering - Concepts - Practices - Perspectives

Modern Physics Simulations

Computer Simulations in Economics

Do Computer Simulations Allow a Better Understanding of Basic Electrical Circuits than Real Lab Experiments?

Simulation in engineering education: The transition from physical experimentation to digital immersive simulated environments

A Study of Computer-Based Simulations for Nano-Systems and their types

Exploration on computer simulation method in physics education

Tools of computer simulation in learning physics

Computer simulation for undergraduate engineering education

Methods for Exploring Simulation Models

Integrative biological simulation praxis: Considerations from physics, philosophy, and data/model curation practices

Data simulations for advancing psychological research: Insights, preparations and investigations

Improving engineering students' understanding of classical physics through visuo-haptic simulations

A Comprehensive Review of Simulation Technology: Development, Methods, Applications, Challenges and Future Trends

Design and Application of Interactive Simulations in Problem-Solving in University-Level Physics Education

A computational mechanics special issue on: data-driven modeling and simulation—theory, methods, and applications

On the feasibility of foundational models for the simulation of physical phenomena

Simulating Physics with Computers

A brief overview of the evolution of computer simulations in economic research

Modeling the social organization of science

A Case Study: Novel Group Interactions through Introductory Computational Physics